Feature Stories | Jun 14 2012
By Greg Peel
Anyone living in Sydney is probably questioning, having just suffered through several days of unseasonably heavy rain, whether recent meteorologist claims that La Nina has now receded are indeed accurate. Sydney may have been in the grips of a Big Wet but any grumbling quickly seems petty when one recalls the devastation wrought across the country by floods and cyclones over the past couple of years.
As far as the local insurance industry is concerned, 2011 was a real shocker. Not only have global interest rates remained low, which impacts on the discounted value of reserves and on returns from premiums invested, the level of catastrophe claims of the past year have skyrocketed above the historical average. Reserves have been wiped out and reinsurance costs have risen.
There appears little hope in the short term of a return to a higher global interest rate environment, and locally the RBA is well into an easing phase. Insurance companies are now lifting premium prices, which is the typical response following a tough claims period, but realistically there is still a bit of “fingers crossed” going on. With reserves depleted, and no investment opportunities offering a quick top-up, insurance companies cannot afford to be hit with further elevated claim levels.
There is, however, a lot of faith being placed in historical cycles. Insurance companies know from experience that while natural catastrophes are unpredictable at best, “cat” levels tend to flow over in regular up-down cycles. The following graph highlights cat claims in dollar value (in today's dollars) going back to 1967. One can quickly note that the intervals between elevated claim periods are fairly regular, and that each peak tends to usher in another period of quiet.
Investors, too, are revisiting the insurance sector. The typical cat cycles plays more importantly into an overlaying earnings cycle, in which the higher premiums applied following a heavy cat period are maintained in ensuing years when cat levels drop, thus raising margins. This thus represents an industry up-cycle. Aside from having faith in these cycles, insurance companies would also be taking comfort from recent meteorologist suggestions that La Nina – bringer of floods and cyclones – is once again drifting away.
But might she make a rapid return?
In February, 2011, I published an article from an objective standpoint highlighting a further weather cycle theory beyond that of El Nino-La-Nina. Noting that the article was written in the wake of the Brisbane floods, but ahead of further weather events throughout 2011 and into 2012, the following is an unedited excerpt of what I wrote at the time:
“So it is from this relatively neutral stance that I make note of a scientific theory that has been gaining some traction of late – that of the Pacific Decadal Oscillation (PDO). I qualify this report straight away by noting the Australian Bureau of Meteorology is so far undecided on the theory's validity but would like to pursue further research before arriving at any opinion.
“We are all now familiar with the incidences of El Nino and La Nina. We understand they are related to periods of warming and cooling in the Pacific and other ocean waters. These cycles can be weak or strong, and the recent drought in Australia coincided with strong El Nino periods and the current 'wet' coincides with a particularly strong La Nina.
“On average, these cycles last six to eighteen months and occur every three to seven years. There are nevertheless no hard and fast rules, and long range prediction is as good as impossible. The best we can do at the moment is see a cycle coming only when it's basically right on top of us, and then monitor when it is the temperature variations begin to turn back again. Meteorologists were able to tell us that a La Nina was apparently beginning late last year but they have only subsequently been able to note that this is a particularly severe cycle.
“But with the benefit of historical records, research and modelling, the theory of the PDO has arisen. This suggests that overlaying the shorter, sharper El Nino and La Nina cycles are longer wet/dry cycles which last 20-30 years. Both the opposing short cycles come and go within each longer cycle, but typically if the longer PDO cycle is 'dry' then El Ninos are more severe and La Ninas are less severe, and vice versa for 'wet' PDO periods.
“Looking at the data for the twentieth century through to today, 1900 fell in a wet PDO which lasted until 1924, a dry PDO occurred from 1925-46, another wet from 1947-76, and a dry from 1977 on. But given the severity of the current La Nina, which has coincided with the breaking of one of Australia's most severe drought periods, the question is: Have we now cycled into the next wet period? The timing is certainly right given a wet period is due. If so, we could be in such a period for another 20-30 years.
“Now let me reiterate – I am not a scientist. But out of curiosity, I thought I would create a table to explore the implications of the PDO for Australia back to 1900. It's lengthy, but the table can be accessed here in Excel format.
“Using data from scientific websites (not Wikipedia) I have created four columns in my table. Column one is each year from 1900 rounded to six-moth intervals. Column two is the periods of PDO, with dry periods represented as red and wet as blue. Column three shows the periods of El Nino (red) and La Nina (blue). Column four shows periods of drought (red) and what the Bureau of Meteorology lists as 'severe' flood incidents (blue).
“The first impression is that the results are not 'perfect'. Droughts have occurred in PDO wet periods and floods in PDO dry periods. There was even one flood (1940) right in the middle of a long drought and El Nino period. According to records, this flood confounded meteorologists at the time but was correctly predicted by aboriginal elders.
“What is striking, however, are what I call the 'triple red' and 'triple blue' periods. Australia's longest droughts have occurred when the PDO is dry and El Nino is occurring. Note the periods 1937-47, 1991-95 and 2000-10 compared to other drought periods. Also note that while all the floods here are noted as 'severe', the most severe floods Australia has experienced prior to 2011 were in 1974 (Brisbane, of which we have all been reminded), 1955 (Hunter Valley, made famous by the movie Newsfront), and 1916 (Clermont Qld, inland from Mackay). Notably the Clermont flood occurred as a result of a cyclone which passed through Townsville. The most severe floods have occurred as 'triple blues', when La Nina has arrived during a PDO wet.
“What one can draw from my table, I believe, were the PDO theory to be granted scientific currency, is that if an El Nino occurs during a PDO dry the chance of severe drought is amplified (but not guaranteed) and if a La Nina occurs during a PDO wet the chance of severe flood is amplified (but not guaranteed).
“As I suggested earlier, scientists are now considering that the strong La Nina that is now upon us, subsequent to the breaking of the long drought, may signal the beginning of a new PDO wet cycle. If so, farmers can rest a little easier about the question of water supply to crops, but may face more episodes of flooding. The Murray-Darling Basin may rejuvenate itself long before politicians come up with a viable solution. Miners may well be in for more regular incidents of lost production from flooded mines.
“I personally am not endorsing anything here – just throwing it subject up for discussion. It is, however, interesting to note that what one might call a 'skeptical' school of scientists (and again I don't mean any on the payroll of Exxon etc) points to the PDO as a possible explanation for global warming beyond that of man-made emissions.
“I also note, again without qualification, that to jump on Australia's recent weather as 'confirmation' of the impact of man-made emissions is to ignore the wider sample set. For example, the 2011 Brisbane flood did not quite reach the height of the 1974 Brisbane flood. Cyclone Yasi did not quite prove more severe than Cyclone Tracy, which also hit in 1974. Back in the seventies scientists were actually worried the earth was cooling, such that a new Ice Age may be upon us.
“Food for thought.”
Food for thought indeed (back live now). I do not wish to revisit any climate change debate in this follow up article but rather I have been reminded of the PDO theory and this earlier article during recent discussions with investors and insurance company representatives, relating to aforementioned faith in a new low-cat cycle (and thus insurance sector earnings up-cycle) being ahead of us.
You'll note that the above cat cycle chart (provided by fund manager Alphinity) highlights four particularly bad cat years, being 1974, 1989, 1999 and 2011. If we take a look at the Excel table I created back in 2011 we can try lining these high cat event years up.
Firstly, 1974 is a clear “triple blue”. The only other “triple blue” within the time frame provided by the cat chart (1967-2011) is 2011. The period 1977 to 2010 is a red period, meaning a longer term PDO “dry” cycle. This cycle broke (assuming it now has) only after 10-20 years of devastating drought. Yet if we look at 1989 we note La Nina was in force and we did indeed see a “severe” flood episode which, on my table, appears as the first half of 1990. And moving down to 1999 we see exactly the same – longer term dry but La Nina and another flood event.
When I wrote the article in 2011 my underlying question was basically: Have we just entered a longer term wet cycle? Over a year later, it sure seems that way, assuming the PDO theory holds water (pardon the pun). We again note that the above cat graph goes back only as far as 1967, just capturing the last few years of the underlying PDO wet. The ensuing period has been one of a PDO dry, at least up to 2011. Now take another look at the Excel table in the years up to 1967 and we note further “triple blues”, albeit there are also flood events within the Long Dry and even within El Nino periods.
As I suggested, there is no exact science here. What I do know, however, is that while no less devastating, droughts themselves do not attract catastrophe insurance payouts. Floods and cyclones certainly do, and if the PDO is accurate and we have now entered a longer term wet cycle, further (inevitable) visits from La Nina could well ensure faith in insurance cycles might be sorely tested.
